Development of scalable methods for the utilization of multi-walled carbon nanotubes in polymer and metal matrix composites

(1)

Dissertations

2014

Development of scalable methods for the

utilization of multi-walled carbon nanotubes in

polymer and metal matrix composites

Danny Vennerberg Iowa State University

Follow this and additional works at:https://lib.dr.iastate.edu/etd Part of theNanoscience and Nanotechnology Commons

This Dissertation is brought to you for free and open access by the Iowa State University Capstones, Theses and Dissertations at Iowa State University Digital Repository. It has been accepted for inclusion in Graduate Theses and Dissertations by an authorized administrator of Iowa State University Digital Repository. For more information, please [email protected].

Recommended Citation

Vennerberg, Danny, "Development of scalable methods for the utilization of multi-walled carbon nanotubes in polymer and metal matrix composites" (2014).Graduate Theses and Dissertations. 13714.

(2)

Development of scalable methods for the utilization of multi-walled carbon nanotubes in polymer and metal matrix composites

by

Daniel Curtis Vennerberg

A dissertation submitted to the graduate faculty in partial fulfillment of the requirements for the degree of

DOCTOR OF PHILOSOPHY

Major: Materials Science and Engineering Program of Study Committee: Michael R. Kessler, Major Professor

Ashraf Bastawros Nicola Bowler Kaitlin Bratlie Ludovico Cademartiri

Iowa State University Ames, Iowa

2014

(3)

LIST OF TABLES

Table 3-1. Elemental surface composition of o-MWCNTs and silanized MWCNTs as

determined by XPS ...64

Table 3-2. Summary of tan δ peak features observed from dynamic mechanical analysis ...70

Table 3-3. Tensile properties of neat epoxy and nanocomposites prepared with 0.1 wt% loading of MWCNTs ...79

Table 3-4. Tensile properties of neat epoxy and nanocomposites prepared with 0.2 wt% loading of MWCNTs ...79

Table 5-1. Literature reports of electrical anisotropy in aligned BP at room temperature ....126

Table 6-1. Vickers microhardness measurements ...145

Table A-1. Naming conventions adopted for the samples investigated in this study ...162

Table A-2. Examples of reaction models ...164

Table A-3. Comparison of model fits for neat epoxy ...166

Table A-4. Summary of kinetic parameters and fit obtained by Prout-Tompkins model for all samples in study. ...168

Table A-5. Summary of kinetic parameters obtained by model free analysis at low levels of conversion for all samples in study. ...171

(8)

LIST OF FIGURES

Figure 1-1. SWCNT, DWCNT, and MWCNT structures (from left to right)[2] ...5

Figure 1-2. Schematic showing how a graphene sheet (a) can be “rolled” up into

zig-zag (b), armchair (c), and chiral (d) nanotubes[5] ...6

Figure 1-3. Experimental setup used by Ruoff to perform tensile tests on individual

CNTs[17] ...11

Figure 1-4. Schematic depiction of noncovalent functionalization via polymer

wrapping[42] ...16

Figure 2-1. Schematic depiction of fluidized bed reactor. ...30

Figure 2-2. Representative SEM and TEM images of pristine n-MWCNTs (a-b) as-received, (c-d) fluidized for 10 min, and (e-f) fluidized for 90 min. Scale bars represent 1 µm and 5 nm for SEM and TEM images, respectively. Red arrows demarcate sidewall functionalization and dashed yellow arrows highlight

amorphous carbon. ...32

Figure 2-3. Representative SEM and TEM images of k-MWCNTs (a-b) as-received, (c-d) fluidized for 10 min, and (e-f) fluidized for 90 min. Scale bars represent 1 µm and 5 nm for SEM and TEM images, respectively. Red arrows demarcate sidewall functionalization. ...33

Figure 2-4. Representative Raman spectra of n-MWCNTs and k-MWCNTs. ...34

Figure 2-5. FTIR spectra of (a) n-MWCNTs and (b) k-MWCNTs ...37

Figure 2-6. XPS survey spectra of (a) n-MWCNTs and (b) k-MWCNTs as a function of O3 exposure time. All spectra are normalized to the C1s peak. ...38

(9)

Figure 2-7. R-values obtained from Raman spectra and surface oxygen concentration of

(a) n-MWCNTs and (b) k-MWCNTs. ...40

Figure 2-8. High resolution XPS C1s spectra of (a) pristine and (b) 90 min O3 n-MWCNTs as well as (c) pristine and (d) 90 min O3 k-MWCNTs. ...42

Figure 2-9. Relative functional group concentrations obtained from deconvolution of high resolution XPS scans for (a) n-MWCNTs and (b) k-MWCNTs as a function of treatment time...44

Figure 2-10. Protic functional group concentrations obtained from Boehm titration of (a) n-MWCNTs and (b) k-MWCNTs. ...45

Figure 3-1. Grafted siloxane structure expected from the reaction of o-MWCNTs with ADMS and ATMS ...57

Figure 3-2. Thermogravimetric analysis of a) ds-MWCNTs and b) ts-MWCNTs ...61

Figure 3-3. FTIR spectra of a) ds-MWCNTs and b) ts-MWCNTs...63

Figure 3-4. XPS survey scans of a) ds-MWCNTs and b) ts-MWCNTs ...65

Figure 3-5. High resolution scans of the Si2p peaks of ds-MWCNTs and ts-MWCNTs reacted under silane concentrations of 0.1, 0.5, and 1.0%. Fits of the Si-O-Si peak ( ), Si-O-C peak ( ), and their sum ( ) are shown superimposed over each raw spectrum ...66

Figure 3-6. TEM images of composites made with 0.2 wt% loading of a) p-MWCNTs, b) ds-MWCNTs, and c) ts-MWCNTs. Insets provide higher magnification views of the dispersion state and structure of MWCNTs in the epoxy. ...68

Figure 3-7. Dynamic mechanical analysis of composites containing 0.1 wt% loading of a) ds- MWCNTs and b) ts-MWCNTs. The main graphs display storage modulus while insets show tan δ. ...71

(10)

Figure 3-8. Dynamic mechanical analysis of composites containing 0.2 wt% loading of a) ds- MWCNTs and b) ts-MWCNTs. The main graphs display storage modulus

while insets show tan δ. ...72

Figure 3-9. CRR of neat epoxy and nanocomposites loaded with 0.2 wt% MWCNTs. ...76

Figure 3-10. Schematic drawing illustrating the proposed microstructure evolution in the neat epoxy with ds-MWCNT and ts-MWCNT additions based on CRR and

DMA measurements. ...77

Figure 3-11. Stress-strain curves of composites made with 0.1 wt% loading of a)

ds-MWCNTs and b) ts-ds-MWCNTs. ...80

Figure 3-12. Stress-strain curves of composites made with 0.2 wt% loading of a)

ds-MWCNTs and b) ts-ds-MWCNTs. ...81

Figure 3-13. SEM images of the fracture surfaces of a) neat epoxy and b) 0.1 wt%

loading of p-MWCNTs. Scale bars represent 10 µm. ...83

Figure 3-14. SEM images of the fracture surfaces of 0.1 wt% loading of a) (0.1)ds-MWCNTs, c) (0.5)ds-(0.1)ds-MWCNTs, and e) (1.0)ds-MWCNTs as well as b) (0.1)ts-MWCNTs, d) (0.5)ts-(0.1)ts-MWCNTs, and f) (1.0)ts-MWCNTs. Scale bars represent 10 µm ...84

Figure 4-1. Reaction of o-MWCNT with GPTMS. ...95

Figure 4-2. Thermogravimetric analysis of p-MWCNTs, o-MWCNTs, and

s-MWCNTs. ...98

Figure 4-3. FTIR spectra of p-MWCNTs, o-MWCNTs, and s-MWCNTs. ...99

Figure 4-4. XPS analysis of MWCNTs; (a) survey scans of p-MWCNTs, o-MWCNTs, and MWCNTs and (b) high resolution spectrum of the Si2p peak from

s-MWCNTs showing fits of the Si-O-Si peak ( ), Si-O-C peak ( ), and their

(11)

Figure 4-5. Rheological behavior of neat resin as well as resin containing 0.5 wt%

p-MWCNTs and s-p-MWCNTs. ...101

Figure 4-6. Dynamic mechanical analysis of nanocomposites made with different

loadings of p-MWCNTs. ...102

Figure 4-7. Dynamic mechanical analysis of nanocomposites made with different

loadings of s-MWCNTs. ...103

Figure 4-8. Changes in the cooperatively rearranging region length with p-MWCNT

and s-MWCNT additions. ...105

Figure 4-9. Tensile properties of neat epoxy as well as p-MWCNTs and

s-MWCNT/epoxy nanocomposites: (a) Young’s modulus, (b) tensile strength, (c)

fracture strain, and (d) toughness ...106

Figure 4-10. Fracture surfaces of (a) neat epoxy and composites filled with (b) 0.1 wt% p-MWCNT and (c) 0.5 wt% p-MWCNT. A higher magnification image of the 0.5 wt% p-MWCNT composite (d) shows significant nanotube pullout. ...108 Figure 4-11. Fracture surfaces of composites filled with (a) 0.1 wt% s-MWCNT and (b)

0.5 wt% s-MWCNT. A higher magnification image of the 0.5 wt% s-MWCNT

composite (c) shows nanotube crack bridging. ...109

Figure 5-1. Schematic of modified Taylor-Couette system used to simultaneously shear and filter MWCNT dispersions ...117

Figure 5-2. Optical image of dried BP sheet formed under a shear rate of 1000 s-1. ...119

Figure 5-3. Rheological behavior of the aqueous MWCNT dispersion used in this

study. ...121

Figure 5-4. Scanning electron micrographs of BP formed at shear rates of 0 s-1, 640 s-1, 825 s-1, and 1000 s-1. ...122

(12)

Figure 5-5. Representative I-V curves for BP samples produced in the absence of shear (red) and at γ 1000 s 1 (blue) measured both parallel and perpendicular to the direction of flow. ...124

Figure 5-6. Summary of electrical conductivity measurements performed on BP in

directions parallel and perpendicular to the direction of flow at various shear rates. ....124

Figure 5-7. Representative stress-strain curves for BP produced in the absence of shear and at γ 1000 s 1 in directions both parallel and perpendicular to flow...126

Figure 5-8. Summary of mechanical properties for BP prepared in the absence of shear and at γ 1000 s 1 in directions both parallel ( ) and perpendicular to flow (

). ...127

Figure 6-1. Schematic depiction of experimental setup ...136

Figure 6-2. Microstructural evolution of the unmixed Cu2O exposed to microwaves in air for 5 s (a), 10 s (b), 30 s (c), and 60 s (d). A higher magnification image of the 30 s sample is given in (e) to better resolve the eutectic structure. A representative micrograph of an etched 60 s sample is shown in (f). ...139

Figure 6-3. XRD spectra of mixed Cu2O samples irradiated in air as a function of

microwave exposure time. ...140

Figure 6-4. Products formed after microwave irradiation of mixed Cu2O in air for 10 s

(a), 30 s (b), and 60 s (c). ...142

Figure 6-5. Products formed after microwave irradiation of Cu2O in argon atmosphere for 15 s in the unmixed (a) and mixed conditions (b) as well as 45 s unmixed (c) and mixed (d). A photograph of the copper pieces formed after microwaving

mixed Cu2O for 30 s (left) and 45 s (right) is shown in (e). ...143

Figure 6-6. Raman spectra of MWCNTs subjected to different irradiation times in air and argon. From bottom to top: pristine MWCNTs, air 30 s, air 45 s, air 60 s, argon 30 s, and argon 45 s. The dashed lines are a guide for the eyes to distinguish

(13)

Figure 6-7. Backscattered SEM image (a) and corresponding EDS maps (b and c) of sample irradiated for 45 s in argon. An overlay of the copper and carbon maps on

the backscattered image is shown in the bottom right figure (d). ...147

Figure 6-8. Secondary electron SEM images of mixed Cu2O irradiated for 45 s in argon. Image (a) was taken from an area appearing to be predominantly copper (light phase in backscattered image), and (b) was captured in an area appearing to be a CNT agglomeration (dark phase in backscattered image). Individual CNT

ends emerging from the copper matrix are indicated with red arrows in (a). ...149

Figure 7-1. MgO nanocubes formed under microwave heating in the presence of MWCNTs for 30 s (a) and higher magnification image of the nanocubes

highlighting their dimensions (b).. ...160

Figure 7-2. AlN nanorods formed by heating Al2O3 for 60 s in the presence of

MWCNTs (a) and and the end of a rod capped with an iron-rich sphere (b). ...160

Figure A-1. Raw DSC curves for neat epoxy ...165

Figure A-2. Degree of cure as a function of temperature and heating rate for neat epoxy ....166

Figure A-3. Model fit of DSC data for the neat epoxy. Solid lines represent raw data

while the colored symbols represent predictions from the Prout-Tompkins model. ...167

Figure A-4. Friedman plots used to determine model-free kinetic parameters as a function of conversion for the neat resin as well as composites containing 0.1 wt% loadings of p-MWCNTs, 0.1ts-MWCNTs, and 1.0ts-MWCNTs ...169

Figure A-5. Dependence of model-free kinetic parameters on conversion for the neat resin as well as composites containing 0.1 wt% loadings of p-MWCNTs,

0.1ts-MWCNTs, and 1.0ts-MWCNTs...170

(14)

Figure A-7. 0.1 wt% p-MWCNT ...174

Figure A-10. 0.1 wt% o-MWCNT ...177

Figure A-11. 0.2 wt% o-MWCNT ...178

Figure A-12. 0.1 wt% 0.1ds-MWCNT ...179 Figure A-13. 0.2 wt% 0.1ds-MWCNT ...180 Figure A-14. 0.1 wt% 1.0ds-MWCNT ...181 Figure A-15. 0.2 wt% 1.0ds-MWCNT ...182 Figure A-16. 0.1 wt% 0.1ts-MWCNT ...183 Figure A-17. 0.2 wt% 0.1ts-MWCNT ...184 Figure A-18. 0.1 wt% 1.0ts-MWCNT ...185 Figure A-19. 0.2 wt% 1.0ts-MWCNT ...186

Figure A-20. 0.1 wt% sc-CO2 MWCNTs ...187

(15)

Figure A-22. 0.5 wt% sc-CO2 MWCNT ...189

Figure B-1. Rheological behavior of neat resin as well as resin containing p-MWCNTs and o-MWCNTs at low and high loading levels. ...192

Figure B-2. Rheological behavior of neat resin as well as resin containing p-MWCNTs and o-MWCNTs at low and high loading levels. ...193

Figure B-3. Rheological behavior of resin loaded with 0.1 wt% ds-MWCNTs. ... 194

Figure B-4. Rheological behavior of resin loaded with 0.5 wt% ds-MWCNTs. ... 195

Figure B-5. Rheological behavior of resin loaded with 0.1 wt% ts-MWCNTs. ... 196

(16)

ACKNOWLEDGEMENTS

I would like express my sincere gratitude to my advisor, Dr. Michael R. Kessler, for his continual support, expertise, and patience throughout my doctoral studies. This dissertation would not have been possible without his insightful guidance to focus my efforts when needed while giving me the latitude to explore new avenues when promising. I would also like to thank Dr. Kaitlin Bratlie, Dr. Ashraf Bastawros, Dr. Nicola Bowler, and Dr. Ludovico Cademartiri for serving on my advisory committee and providing both technical and personal advice.

This work was supported by a National Science Foundation Graduate Research Fellowship, and I am deeply grateful to the federal government for the generous support. Additional funding was provided by Kumho Petrochemical Co. Correspondence with Kumho’s technical team helped to ensure the scalability of the research and broadened my graduate experience.

I would also like to thank Jim Anderegg of the Ames Laboratory for his generous help in characterizing many of my samples and for the enlightening discussions we had. My time at Iowa State was made enjoyable and productive in large part to the current and former members of the Polymer Composites Research Group. I thank Mitch Rock, Dr Rafael Quirino, Dr. Samy Madbouly, Rui Ding, Amy Bauer, Hongchao Wu, Ruqi Chen, Dr. Eliseo De Leon, Dr. Peter Hondred,, Dr. Tom Garrison, Chaoqun Zhang, Hongyu Cui, Kunwei Liu, Gauri Ramasubramanian, and Mike Zenner for their friendship and technical assistance. Zach Rueger and Ryan Hall deserve special thanks for their help in preparing and characterizing composite samples.

(17)

I want to thank my family for their unwavering support throughout my academic career. My accomplishments are only possible because of the love and encouragement of my parents, for whom I am truly grateful. Finally, I want to thank my fiancé, Kelsey, for her patience and support throughout my graduate studies.

(18)

ABSTRACT

Multi-walled carbon nanotubes (MWCNTs) have received considerable attention as reinforcement for composites due to their high tensile strength, stiffness, electrical conductivity and thermal conductivity as well as their low coefficient of thermal expansion. However, despite the availability of huge quantities of low-cost, commercially synthesized nanotubes, the utilization of MWCNTs in engineering composites is extremely limited due to difficulties in achieving uniform dispersion and strong interfacial bonding with the matrix. A proven method of enhancing the nanotube-polymer interface and degree of MWCNT dispersion involves functionalizing the MWCNTs through oxidation with strong acids. While effective at laboratory scales, this technique is not well-suited for large-scale operations due to long processing times, poor yield, safety hazards, and environmental concerns.

This work aims to find scalable solutions to several of the challenges associated with the fabrication of MWCNT-reinforced composites. For polymer matrix composite applications, a rapid, dry, and cost-effective method of oxidizing MWCNTs with O3 in a fluidized bed was developed as an alternative to acid oxidation. Oxidized MWCNTs were further functionalized with silane coupling agents using water and supercritical carbon dioxide as solvents in order to endow the MWCNTs with matrix-specific functionalities. The effect of silanization on the cure kinetics, rheological behavior, and thermo-mechanical properties of model epoxy nanocomposites were investigated. Small additions of functionalized MWCNTs were found to increase the glass transition temperature, strength, and toughness of the epoxy.

In order to achieve composite properties approaching those of individual nanotubes, new approaches are needed to allow for high loadings of MWCNTs. One strategy involves

(19)

making macroscopic mats of nanotubes called buckypaper (BP) and subsequently infiltrating the mats with resin in processes familiar to traditional fiber-reinforced composites. The latter part of this thesis work explores a new method of producing BP comprised of oriented nanotubes through the use of a modified Taylor-Couette setup capable of simultaneously shearing and filtering an aqueous MWCNT dispersion. BP produced with this setup exhibited anisotropic electrical and mechanical properties as a result of the nanotube alignment.

Finally, a new technique for producing MWCNT metal matrix composites was developed using the nanotubes as the heating element and carbon source in a microwave-assisted carbothermic reduction of copper oxide. The extremely rapid heating of MWCNTs upon microwave irradiation allowed Cu-MWCNT composites to be produced in times on the order of a minute. Because this approach requires none of the specialized equipment generally used in metal matrix composite processing, it has promise as a scalable fabrication technique.

(20)

CHAPTER 1:

GENERAL INTRODUCTION

1.1 Introduction

The discovery of carbon nanotubes (CNTs) and their extraordinary mechanical, thermal, and electrical properties has sparked significant effort to develop lightweight composites using CNTs as reinforcement. While individual nanotubes have properties at the extreme of known materials, composites filled with CNTs often fall short of expected improvements due to poor dispersion and interfacial bonding. Over the last two decades, a significant amount of research has been devoted to modifying the surface of multi-walled carbon nanotubes (MWCNTs) for use in polymer matrix composites, and several promising approaches have been discovered. Among these, covalent functionalization of the outer walls of MWCNTs has proven effective at improving nanotube dispersion in and interfacial bonding with polymer matrices. Unfortunately, the vast majority of functionalization reactions are not amenable to industrial use because of long processing times, poor yield, safety hazards, and environmental concerns. Recent commercial expansion in the synthesis of MWCNTs has enabled the production of high quality nanotubes in huge quantities at low cost and opened the door for widespread application of MWCNT-reinforced composites. However, in order for the industrial production of these composites to become viable, alternative functionalization strategies are needed to address the unique challenges of nanotube modification on a large scale.

This thesis encompasses my efforts to develop scalable strategies for covalently functionalizing MWCNTs and subsequently using them to fabricate polymer matrix composites. A two-step reaction is adopted to functionalize the MWCNTs, which consists of first oxidizing the nanotubes and subsequently using the oxygen-bearing moieties as a

(21)

platform for further functionalization with silane coupling agents. This approach deliberately allows for a great deal of flexibility in the final functionality of the nanotube, as silane coupling agents with a wide variety of terminal functional groups are commercially available at low cost. While the functionalized MWCNTs can be made compatible with many resin systems, the scope of this work is limited to the incorporation of nanotubes in a model epoxy. The effects of both pristine and functionalized MWCNT additions on the thermo-mechanical, rheological, and curing behaviors of the epoxy are evaluated. Economic, environmental, and regulatory constraints associated with large-scale nanocomposite production are considered at each stage of composite fabrication and silanization and composite processing are performed without the use of any organic solvents.

While CNT composite research has been dominated by polymer matrices, metal matrix composites reinforced with CNTs also have great potential because the same advantages CNTs afford polymer matrix composites apply to metal matrix composites as well. In particular, CNTs exhibit higher strength, stiffness, thermal conductivity, and electrical conductivity than the matrix phase while having lower density. A major reason that CNT-metal matrix composites are not as well established as PMCs is due to the fact that metals are more difficult to process than polymers, because high temperatures and/or large forces are usually needed to introduce reinforcement into the metal matrix. In addition to work on polymer nanocomposites, this dissertation also presents a novel method of producing carbon nanotube metal matrix composites. Using a microwave-assisted carbothermic route, copper-MWCNT composites are prepared in very short times (on the order of a minute) starting from copper oxide and pristine MWCNTs. This new approach for fabricating carbon nanotube-reinforced metal matrix composites eliminates many of the

(22)

challenges associated with traditional methods while requiring a fraction of the time and energy.

1.2 Dissertation Organization

This dissertation is organized into eight chapters with the middle six consisting of manuscripts that have either been published in or prepared for submission to peer-reviewed journals. Chapter 1 serves as in introduction to CNTs and previous efforts to fabricate CNT-reinforced composites. An emphasis is given to the functionalization and dispersion of CNTs in polymer resins, as the bulk of this work focuses on strategies for producing polymer nanocomposites. Chapter 2 discusses a method for the uniform oxidation of MWCNTs with O3 through the use of a fluidized bed and examines the role of nanotube defect structures on their oxidation.

Chapters 3 and 4 encompass efforts to silanize MWCNTs that had been previously oxidized by O3 in the fluidized bed described in Chapter 2. Chapter 3 focuses on aqueous silanization of the MWCNTs using both di-functional and tri-functional amino-silanes. The presence of water in this reaction allowed for the self-condensation of the silane molecules, which resulted in the grafting of two different oligomeric structures on the surface of the nanotube. Chapter 4 focuses on silanization of the MWCNTs using an epoxy-terminated silane with supercritical carbon dioxide as the solvent for the reaction. In contrast to the aqueous route, the absence of water produces a nearly monolayer deposition of silane on the nanotube surface. Both chapters also provide detailed results of the characterization of the thermo-mechanical properties of composites made with the functionalized MWCNTs. An emphasis is placed on understanding the role of surface structure in developing the observed properties.

(23)

Chapter 5 details the development of a novel method to orient MWCNTs in solution and deposit them as a macroscopic sheet termed buckypaper. This approach lends itself to the fabrication of composites with much higher loadings of nanotubes than the types of composites explored in Chapters 3 and 4. Alignment of the nanotubes within the buckypaper sheet produces anisotropic properties, which are useful for composite fabrication and other potential applications such as in sensors or filters.

Chapter 6 describes a completely new method of making metal matrix composites starting from metal oxides using a microwave-assisted carbothermic reaction. The technique is applied to copper oxide to form copper-MWCNT composites with excellent hardness and offers several benefits over traditional metal matrix composite fabrication technologies in terms of safety and ease of processing.

Chapter 7 summarizes the conclusions drawn from the entire body of work. Ideas for future areas of inquiry and extensions of the presented findings are also given. As a supplement to Chapters 3 and 4, appendices A and B review the effects of pristine, oxidized, and silanized MWCNTs on the epoxy curing and rheological behavior, respectively.

1.3 Literature Review 1.3.1 Carbon nanotube structure

CNTs consist of carbon atoms bonded trigonally in a curved, hollow cylinder resembling a sheet of graphene rolled into a tube. These tubes are generally 1-50 nanometers in diameter and microns to millimeters in length, although lengths up to 18.5 cm have been observed.[1] CNT structures are often categorized by the number of concentric graphitic layers making up the tube. Single-walled carbon nanotubes (SWCNTs) consist of only one nanotube, double-walled carbon nanotubes (DWCNTS) are made up of two nested

(24)

nanotubes, and multi-walled carbon nanotubes (MWCNTs) contain several concentric nanotubes. Examples of each type of CNT are shown in Figure 1-1.

Figure 1-1. SWCNT, DWCNT, and MWCNT structures (from left to right)[2]

In addition to the number of nested nanotubes, the CNT structure is also defined by the chirality of each nanotube. The hexagonal rings making up the sidewall of a nanotube have a discrete number of configurations relative to the longitudinal axis of the CNT, which gives rise to a finite set of structural possibilities. The chirality is most easily described by a pair of indices (n,m) defined by a vector called the chiral vector Ch = na1 + ma2, in which n

and m are integers and a1 and a2 are unit vectors shown in Figure 1-2. Creation of a nanotube

from a graphene plane such as that shown in Figure 1-2 can be visualized by rolling the sheet in the direction of the chiral vector n atoms in the a1 direction and m atoms in the a2

direction. The resulting nanotube axis will then be orthogonal to the chiral vector. Using this convention, any SWCNT can be fully defined by simply specifying n and m. If m = 0, the SWCNT is called zig-zag. If n = m, the nanotube is termed armchair. All other configurations are called chiral. The chirality of a nanotube has a profound effect on its electrical properties. If n-m is divisible by 3, the nanotube will be a metallic conductor. Otherwise, it will behave as a semiconductor with a bandgap of ~ 0.5 eV.[3] The concentric

(25)

nanotubes within a MWCNT have a random chirality distribution under most growth conditions, although it is possible to produce monochiral MWCNTs.[4]

Figure 1-2. Schematic showing how a graphene sheet (a) can be “rolled” up into zig-zag (b), armchair (c), and chiral (d) nanotubes[5]

The history of the discovery of CNTs is an interesting and often misrepresented tale. Sumio Ijima is widely credited as the first researcher to produce a CNT after his transmission electron micrographs were published in Nature in 1991.[6] However, CNTs have been synthesized by humans as far back as 1889 from thermal decomposition of methane during attempts by Thomas Edison to produce carbon filaments for light bulbs,[7] although the direct observation of CNTs was not possible until the advent of transmission electron

(26)

microscopy in the 1930s. The first known image of a MWCNT was published in 1952 by Russian scientists,[8] and several other reports of nanotube structures were reported up until 1991.[9] Nevertheless, the report by Ijima was widely read and was published at a time when researchers across many disciplines were eager to explore nano-scale materials. Subsequent work yielded a variety of methods to grow CNTs using summarized in the following section.

1.3.2 Synthesis of CNTs

The three major methods of CNT synthesis today can be classified as arc discharge, visible light vaporization, and chemical vapor deposition (CVD). A catalyst is required to produce SWCNTs with all three routes and for CVD of MWCNTs. At present, CVD is the most economical method of CNT production, and all MWCNTs used in this work were produced by this technique.

Arc discharge represents a relatively simple technique for producing CNTs. The setup usually consists of two moveable graphite rods contained in an air-tight chamber. After evacuation of the chamber, helium or argon gas is flowed continuously until a pressure of ~ 500 Torr is reached. A bias of ~20 V is placed across the graphite rods, which are subsequently brought closer together until current arcs between them. The temperature at the anode rod reaches several thousand degrees Celsius, which causes the carbon to sublimate and eventually condense on the surface of the cathode in the form of MWCNTs and other products such as graphite and amorphous carbon. [10] Metal catalyst can also be introduced into the center of the anode to produce SWCNTs or to control the growth of MWCNTs. The exact mechanism of MWCNT formation by arc discharge is debatable, but SWCNT formation with a catalyst is believed to follow a vapor-liquid-solid (VLS) model regardless of the synthesis route. Arc discharge has the advantage of being relatively easy and

(27)

inexpensive to set up, especially on a lab-scale, and an arc welder can be used to supply current across the electrodes. However, the method has poor yield (generally producing less than 10% nanotube material) and is not easily scaled, as increasing the diameter of the electrodes reduces the fraction of nanotubes produced.

Visible light vaporization involves illuminating a piece of graphite with a high energy light source while under a controlled atmosphere of helium or argon. The graphite becomes hot enough to vaporize and small clusters of gaseous carbon recombine into CNTs given a proper temperature gradient, which is supplied by a cold finger and continuous flow of gas.[11] Several variants of the technique exist that can be defined by their light source: a pulsed laser is used in “laser ablation synthesis”, a continuous laser for “laser vaporization”, or multiple continuous wavelengths from a solar furnace may be used for “solar vaporization”. Visible light vaporization apparatuses are generally more expensive and complicated to set up than arc discharge growth chambers, although they produce better yields of CNTs (up to 50%) and the nanotubes have fewer defects.[11]

CVD represents the third, and most widely used technique for synthesizing CNTs. CVD requires a carbon-bearing gas feedstock and a catalyst.[12] The catalyst, usually a transition metal, lowers the temperature required to thermally decompose a feed gas, which can be a variety of hydrocarbons or CO. The catalyst can be suspended in a gas (floating) or placed on an inert support material such as silica or alumina. In either case, during the synthesis reaction, feed gas is heated until it decomposes (generally 600-1200 °C) into carbon and another byproduct. The carbon dissolves in the catalyst and precipitates out as a CNT according to the VLS model. The key to obtaining high quality CNTs at high yields lies in the design of catalyst particles and in optimizing the furnace atmosphere and temperature.

(28)

Metal catalysts must be small, have a high solubility for carbon, and be stable at high temperatures to prevent coalescence. Ni, Fe, and Co are common constituents in a catalyst formulation and refractory metals are often added to increase the melting temperature; however, the specifics of modern catalyst design are the closely guarded property of CNT manufacturers. The reaction temperature is another critical parameter that must be high enough to decompose the feed gas but low enough to prevent coalescence of the catalyst particles. Thus, different feed gases such as CO, methane, acetylene, and benzene require different reaction conditions and allow for flexibility in developing different catalysts. CVD is the most economically favorable and scalable CNT synthesis technique known.

One of the exciting aspects of CVD for industrial production of CNTs is the fact that the feedstock is very inexpensive (and almost negligible in the case of CO). This situation is the opposite case of polymers, in which monomers dictate the price of the final product. If catalyst efficiency can be improved, CNTs could be produced at very low costs approaching that of the energy needed to heat the reactor. As an illustration, catalyst costs for CNT production are comparable to that of high density polyethylene (HDPE). HDPE yields from ethylene are about 10,000:1 in terms of polymer weight:catalyst weight, and HDPE sells for about $1/lb. Current yields for MWCNTs and SWCNTs are around 10:1 and 0.1:1 and cost ~$100/lb and $10,000/lb, respectively.[11] Thus, yield is the determining factor in the price of CNTs from CVD today, and better catalysts could allow the production of CNTs near or below the price of commodity plastics.

1.3.3 Properties of individual nanotubes

Measuring the properties of an individual CNT is a difficult task not only because nanotubes are so small in dimensions, but also because the properties are affected by the

(29)

types and quantities of defects present, the chirality of the nanotube, and the number of nested layers (in the case of MWCNTs). Nevertheless, serious effort has been expended to determine the mechanical, electrical, and thermal properties of both SWCNTs and MWCNTs and a survey of the findings is given below.

The mechanical properties of CNTs originate from the strong, in-plane sigma bonds formed between sp2-hybridized carbon atoms in the hexagonal sidewall, which have bond energies of ~ 346 kJ/mol, and the weak pi bonding between shells (in MWCNTs) with energies of ~ 3.4 kJ/mol.[13] The modulus of a material is defined by the following equation

1

in which E is tensile modulus, V is volume, G is Gibbs free energy, ε is strain, and σ is stress. One problem with the determination of the modulus (and tensile strength) of nanotubes arises from the definition of the area (A) upon which a force (F) is applied in the / term. One popular convention has been to consider the nanotube 0.34nm thick (the intershell spacing in a MWCNT) and to calculate the area based on the nanotube radius. Using this approach the theoretical modulus of CNTs has been found to be ~ 1 TPa.[14, 15] Experimental characterization of MWCNTs and SWCNTs has yielded moduli on the same order of magnitude.[16-20] The most direct of these measurements involved fixing individual nanotubes between two AFM tips and straining to failure while monitoring with SEM, as shown in Figure 1-3. MWCNTs tested using this setup were found to have moduli values between 0.27–0.95 TPa, failure strains up to 12%, and ultimate tensile strengths from 11-63 GPa.[17] A similar set of measurements on SWCNT bundles yielded modulus values in the

(30)

range of 0.32-1.47 GPa, failure strains on the order of 5.3%, and ultimate tensile strengths from 13-52 GPa.[16]

Figure 1-3.Experimental setup used by Ruoff to perform tensile tests on individual CNTs[17]

CNTs also have extraordinary and complicated electrical properties. Electron flow is confined to the long axis of the nanotube resulting in one-dimensional electronic conduction. As previously mentioned, the chirality of a CNT dictates its electrical behavior. Nanotubes having structures satisfying the condition of (n-m)/3 = integer are metallic and the remainder are semiconductors. Thus, 1/3 of all CNTs have no band gap and the remaining 2/3 are semiconductors with band gaps that are inversely proportional to the tube diameter.[21] This size-dependent behavior arises because curvature in the sidewall slightly misaligns neighboring π orbitals (and may induce partial σ- π hybridization in small diameter tubes), which increases the band gap.[22, 23] For all metallic nanotubes besides armchair structures,

(31)

curvature effects also cause a tiny band gap to form; however, thermal fluctuations at temperatures close to room temperature render these nanotubes metallic. Armchair CNTs are ballistic conductors, meaning that the mean free path of electronic scattering is much larger than the nanotube dimensions. As a result, resistance does not depend on the nanotube length, and no resistance heating is observed because no electron scattering occurs (which normally dissipates heat in the form of phonons via inelastic interactions). Measurements performed on a single 4 mm long, 1.7 nm diameter metallic SWCNT yielded a conductivity of about 5 x 107 S/m.[24] Measurements of MWCNT conductivity are more challenging as a result of multiple shells with varied chirality, and a range of conductivities from 2 x 107 – 8 x 105 have been reported.[25, 26]

Thermal conductivity represents another remarkable property of CNTs. Values of 1400 W/(mK),[27] 2000 W/(mK),[28] and 3000 W/(mK)[29] have been determined from measurements performed on single MWCNTs. The thermal conductivity was found to depend heavily on temperature, reaching a maximum value around 47 °C and decreasing sharply at higher temperatures.[29] Other measurements of individual MWCNTs have yielded lower conductivity values around 300-800 W/(mK).[30, 31] Bundled MWCNTs and buckypaper sheets were found to have even lower thermal conductivities of 150 W/(mK) and 50 W/(mK),[32] respectively, due to thermal resistances caused by inter-tube junctions. The large range in measured values may be due to the fact that the thermal conductivity is theoretically predicted to depend on nanotube length,[33] diameter,[34] and structural defect density[32] (which is relatively high in CVD-synthesized nanotubes) as well as interactions with substrates. The studies performed to date have not explicitly taken account of all of these factors, and, as a result cannot be readily compared.

(32)

1.3.4 Dispersion of CNTs in a matrix

A significant body of work has been devoted to producing CNT-reinforced polymer nanocomposites for structural and functional applications.[35-42] However, despite over two decades of research, CNT composites have still underperformed their potential because of the challenges associated with dispersion of CNTs in the resin during processing and the poor interfacial bonding between nanotubes and matrix. The extremely high aspect ratio and surface area of CNTs makes them more difficult to disperse than traditional fillers or spherical nano-scale particles. Furthermore, commercially available CNTs consist of intertwined bundles that are especially difficult to disentangle. This section will briefly review the most popular techniques that have been adopted to disperse CNTs in polymers.

The most common dispersion method is ultrasonication, which involves disentanglement through the use of ultrasound energy. When CNTs mixed in a resin or solvent are subjected to a high frequency, oscillating pressure wave, shockwaves are produced in the dispersing medium which causes nanotubes on the outer extremities of an agglomerate to separate from the bundle.[43] Ultrasonication requires that the CNTs be dispersed in a low-viscosity liquid in order to limit attenuation. As a result, CNT/resin mixtures are often diluted with an organic solvent during sonication processing; however, sonication produces a substantial amount of heat, and volatile solvents must be actively cooled to prevent loss via evaporation. Two types of sonication equipment are commonly employed: bath and horn sonicators. Bath sonicators generate 15-400 kHz waves through a medium of water at powers less than 100 W. With this technique, CNT mixtures in a container such a glass vial are placed in the water bath and subjected to uniform pressure waves. The relatively mild pressure waves generated by bath sonicators are effective at

(33)

dispersing CNTs in very low viscosity liquids (especially with the aid of surfactants) without severely damaging nanotubes. Horn (or probe) sonicators produce a high powered (100-1500 W), directional pressure waves. While effective at dispersing CNTs, long exposure times can lead to structural damage of the nanotube sidewalls, which degrades electrical and mechanical properties.[44, 45]

Another common method of dispersion is calendaring. With this technique, very viscous resin containing CNTs is fed through a three-roll mill consisting of closely spaced, adjacent rollers that are spinning in alternating directions and varying speeds. As the nanotube-filled resin passes through the small gaps between rollers, it experiences high shear stresses, which break up agglomerations. By controlling the angular velocity of the rollers and the gap size, the magnitude of the shear can be adjusted, and dispersions are often calendered several times using consecutively smaller gap widths. While this method is simple and convenient for homogenizing CNTs, it has limitations. The CNT/resin mixture must be paste-like in viscosity for the rollers to pull the dispersion through the gaps, which restricts its use with most thermosetting resins and some thermoplastics. Furthermore, most three-roll mills have minimum gap sizes of 5 μm, which is insufficient to fully disentangle CNT bundles.[46] Instead, calendaring often distributes microscopic agglomerates of CNTs very evenly in a resin.[46-50]

A third popular dispersion technique involves shear mixing with high-speed mixers or extruders. Laboratory-scale overhead mixers capable of reaching speeds of several thousand rpm have been used to produce fine levels of dispersion in thermosetting resins.[51] However, overhead mixing is restricted to use with low and medium viscosity resins, and reagglomeration of CNTs post-mixing is common. CNTs can be compounded with high

(34)

viscosity resins using extruders. Extrusion is the most popular method of fabricating thermoplastic nanocomposites and the widespread use of industrial extrusion makes this technique readily scalable for engineering applications. Typical processing involves feeding solid plastic pellets mixed with CNTs into an extruder and subjecting them to large shear forces provided by rotating screws, which disperse the CNTs and generate heat to melt the polymer. Melt mixing via extrusion is generally less effective at dispersing the CNTs on a nano-scale level than solvent-assisted sonication due to the lower shear rates involved and the absence of small molecules to infiltrate the interstices between nanotubes in a bundle.[37] Interestingly, while MWCNTs can be adequately dispersed with twin-screw extrusion, the technique has proven less effective for SWCNT/resin mixtures.[11] Nevertheless, extrusion has several advantages over other dispersion methods including simplicity, availability, and the fact that reagglomeration of CNTs is limited by the long relaxation times of viscous polymers.

1.3.5 Functionalization of CNTs

All of the dispersion techniques outlined above can be aided by the introduction of functional groups on the CNT sidewalls that are compatible with the matrix resin. Furthermore, for nanocomposites to attain excellent mechanical properties requires more than simply achieving a homogeneous physical distribution; it requires strong adhesion between the MWCNTs and matrix. Functionalization of the MWCNT surface is a widely acknowledged strategy for improving dispersion and strengthening the interface simultaneously, and previous approaches can generally be divided into two categories: covalent and noncovalent functionalization. A few examples from each category are briefly described below.

(35)

Noncovalent functionalization involves covering the MWCNT surface with polymers [52] or surfactants [53], which are held to the nanotube surface by Van der Waals forces, as shown in Figure 1-4. An advantage of this approach is that the sidewalls of the MWCNT remain intact, so the electrical conductivity of the nanotubes remains very high, and composites with excellent electrical properties have been fabricated.[42] However, because the polymers or surfactants are only weakly attached to the MWCNTs, they easily slip off the nanotube surface under applied stress. Furthermore, many techniques use polymers that are cost-prohibitive for industrial processing.[54, 55]

Figure 1-4.Schematic depiction of noncovalent functionalization via polymer wrapping[42]

Covalent functionalization involves chemical attachment of a moiety to the outer sidewalls or ends of the MWCNTs. Myriad techniques have been explored to functionalize the MWCNT sidewall. Several popular routes have been broadly categorized below as addition reactions, acid oxidation, and physical oxidation.

Like all fullerenes, the MWCNT walls have an associated strain energy which renders them more reactive than planar aromatic molecules. As a result, a variety of chemical addition reactions are possible including fluorination, hydrogenation, and cycloaddition of groups such as carbenes, nitrenes, and azomethines.[42, 56, 57] While this approach allows a

(36)

high degree of control over the nature of functional groups added to the MWCNT, the rigorous reaction conditions and long treatment times required are impractical for industrial-scale processing.

Exposure of MWCNTs to strong acids is a popular method to generate defect sites on the MWCNT surface containing a mixture of functionalities including carboxylic, carbonyl, and hydroxyl groups.[56, 57] Sulfuric acid, sulfuric and nitric acid mixtures, and piranha solution (sulfuric acid and hydrogen peroxide mixture) are commonly used to oxidize MWCNTs,[58] and ultrasonication is often employed to disperse MWCNTs in the acids and improve their oxidation efficiency.[59] This method is simple and effective at introducing oxygen functional groups on the MWCNT surface, but it has several disadvantages. Acid oxidation often has poor yield, and the MWCNT structure is often severely damaged after functionalization, which dramatically decreases the mechanical properties of the nanotubes and reduces their aspect ratios. In addition, the use of strong acids is unattractive for large-scale processing due to safety and environmental concerns.[60]

Physical oxidation involves oxidation of preexisting defects on the outer walls of MWCNTs using sources such as plasmas [61] or UV light.[62, 63] Felten et al. effectively used O2 plasma to oxidize the MWCNT surface and control the nature of functional groups (hydroxyl, carboxylic acid, and carbonyl) by tuning the reaction atmosphere. Asano et al. introduced 3 wt% COOH groups onto MWCNT surfaces after only 1 min of exposure to intense UV light. Physical oxidation methods such as these are industrially attractive because they are nonpolluting, require no harsh chemicals, and can be performed rapidly. However, dry processing in the absence of solvents introduces challenges in attaining uniform MWCNT functionalization.

(37)

1.4 Research Objectives

While many functionalization routes developed in the past have proven effective, very few are amenable to industrial-scale processing, and those that are promising for large-scale production have not been optimized. Furthermore, few studies have addressed the practical challenges of functionalizing huge quantities of CNTs. For instance, any functionalization methods involving the use of solvents generates liquid waste containing CNTs. Proper disposal of such waste provides a costly challenge and may be subject to regulatory scrutiny for environmental and health reasons. Nonetheless, almost all reported functionalization methods employ copious amounts of solvents. This work aims to address the major issues limiting MWCNT utilization on a large scale by developing functionalization routes that use inexpensive and commercially available reagents, do not require organic solvents, and proceed quickly. The great diversity of resin-MWCNT combinations precludes the study of even a fraction of all possible composite systems. The bulk of this work focuses on a model epoxy to test the efficacy of MWCNT functionalization in improving composite properties. However, the techniques reported here have been made intentionally versatile to broaden their utility for any composite application.

The specific goals of this research are to:

(i) Develop dry MWCNT oxidation strategies for rapid, cost effective production of oxidized MWCNTs (o-MWCNTs).

(ii) Functionalize the o-MWCNTs with a variety of silane coupling agents to attach epoxy or amine end groups and facilitate chemical reactions of these

(38)

functionalized MWCNTS (f-MWCNTs) with either the epoxy matrix or the amine hardener.

(iii) Validate the efficacy of the hybrid MWCNT functionalization strategies by preparing epoxy nanocomposites with various loadings of f-MWCNTs

(iv) Characterize the effect of f-MWCNT type and loading on the cure, rheological, thermal, and mechanical properties of the model epoxy matrix to demonstrate the ability of f-MWCNT end groups to react with the epoxy matrix or the hardener

(v) Advance new processing methods that allow MWCNT composites to be fabricated with much higher nanotube loading levels than traditional melt/resin mixing.

(vi) Explore alternative uses for inexpensive, commercially synthesized MWCNTs such as heating media and carbon sources in carbothermic reactions

1.5 References

[1] Wang X, Li Q, Xie J, Jin Z, Wang J, Li Y, et al. Fabrication of Ultralong and Electrically Uniform Single-Walled Carbon Nanotubes on Clean Substrates. Nano Letters. 2009;9(9):3137-41.

[2] Dumé B. Scientists delve deeper into carbon nanotubes. Physics World: Institute of Physics 2013.

[3] Wilder JWG, Venema LC, Rinzler AG, Smalley RE, Dekker C. Electronic structure of atomically resolved carbon nanotubes. Nature. 1998;391(6662):59-62.

(39)

[4] Xu Z, Bai X, Wang ZL, Wang E. Multiwall Carbon Nanotubes Made of Monochirality Graphite Shells. Journal of the American Chemical Society. 2006;128(4):1052-3.

[5] Rao CNR, Voggu R, Govindaraj A. Selective generation of single-walled carbon nanotubes with metallic, semiconducting and other unique electronic properties. Nanoscale. 2009;1(1):96-105.

[6] Iijima S. Helical microtubules of graphitic carbon. Nature. 1991;354(6348):56-8. [7] Bacon R, Bowman J. Production and properties of graphite whiskers. Bulletin of the American Chemical Society. 1957;2:131.

[8] Radushkevich L, Lukyanovich V. strukture ugleroda, obrazujucegosja pri termiceskom razlozenii okisi ugleroda na zeleznom kontakte. Zurn Fisic Chim. 1952;26:88-95.

[9] Monthioux M, Kuznetsov VL. Who should be given the credit for the discovery of carbon nanotubes? Carbon. 2006;44(9):1621-3.

[10] Kar KK. Carbon Nanotubes: Synthesis, Characterization and Applications: Research Pub.

[11] Grady BP. Carbon Nanotube-Polymer Composites: Manufacture, Properties, and Applications: Wiley; 2011.

[12] Kumar M, Ando Y. Chemical Vapor Deposition of Carbon Nanotubes: A Review on Growth Mechanism and Mass Production. Journal of Nanoscience and Nanotechnology. 2010;10(6):3739-58.

[13] Mukhopadhyay SM, Karumuri A, Barney IT. Hierarchical nanostructures by nanotube grafting on porous cellular surfaces. Journal of Physics D: Applied Physics. 2009;42(19):195503.

[14] Lu JP. Elastic Properties of Carbon Nanotubes and Nanoropes. Physical Review Letters. 1997;79(7):1297-300.

[15] Hernández E, Goze C, Bernier P, Rubio A. Elastic properties of single-wall nanotubes. Appl Phys A. 1999;68(3):287-92.

[16] Yu M-F, Files BS, Arepalli S, Ruoff RS. Tensile Loading of Ropes of Single Wall Carbon Nanotubes and their Mechanical Properties. Physical Review Letters. 2000;84(24):5552-5.

(40)

[17] Yu M-F, Lourie O, Dyer MJ, Moloni K, Kelly TF, Ruoff RS. Strength and Breaking Mechanism of Multiwalled Carbon Nanotubes Under Tensile Load. Science. 2000;287(5453):637-40.

[18] Poncharal P, Wang ZL, Ugarte D, de Heer WA. Electrostatic Deflections and Electromechanical Resonances of Carbon Nanotubes. Science. 1999;283(5407):1513-6. [19] Salvetat J-P, Briggs GAD, Bonard J-M, Bacsa RR, Kulik AJ, Stöckli T, et al. Elastic and Shear Moduli of Single-Walled Carbon Nanotube Ropes. Physical Review Letters. 1999;82(5):944-7.

[20] Treacy M, Ebbesen T, Gibson J. Exceptionally high Young's modulus observed for individual carbon nanotubes. 1996.

[21] Loiseau A. Understanding Carbon Nanotubes: From Basics to Applications: Springer; 2006.

[22] Ding J, Yan X, Cao J, Wang D, Tang Y, Yang Q. Curvature and strain effects on electronic properties of single-wall carbon nanotubes. Journal of Physics: Condensed Matter. 2003;15(27):L439.

[23] Popov VN. Curvature effects on the structural, electronic and optical properties of isolated single-walled carbon nanotubes within a symmetry-adapted non-orthogonal tight-binding model. New Journal of Physics. 2004;6(1):17.

[24] Mann D, Javey A, Kong J, Wang Q, Dai H. Ballistic Transport in Metallic Nanotubes with Reliable Pd Ohmic Contacts. Nano Letters. 2003;3(11):1541-4.

[25] Dai H, Wong EW, Lieber CM. Probing Electrical Transport in Nanomaterials: Conductivity of Individual Carbon Nanotubes. Science. 1996;272(5261):523-6.

[26] Ebbesen TW, Lezec HJ, Hiura H, Bennett JW, Ghaemi HF, Thio T. Electrical conductivity of individual carbon nanotubes. Nature. 1996;382(6586):54-6.

[27] Li Q, Liu C, Wang X, Fan S. Measuring the thermal conductivity of individual carbon nanotubes by the Raman shift method. Nanotechnology. 2009;20(14):145702.

[28] Fujii M, Zhang X, Xie H, Ago H, Takahashi K, Ikuta T, et al. Measuring the Thermal Conductivity of a Single Carbon Nanotube. Physical Review Letters. 2005;95(6):065502. [29] Kim P, Shi L, Majumdar A, McEuen PL. Thermal Transport Measurements of Individual Multiwalled Nanotubes. Physical Review Letters. 2001;87(21):215502.

[30] Choi T-Y, Poulikakos D, Tharian J, Sennhauser U. Measurement of the Thermal Conductivity of Individual Carbon Nanotubes by the Four-Point Three-ω Method. Nano Letters. 2006;6(8):1589-93.

(41)

[31] Choi TY, Poulikakos D, Tharian J, Sennhauser U. Measurement of thermal conductivity of individual multiwalled carbon nanotubes by the <equation>3-ω</equation> method. Applied Physics Letters. 2005;87(1):013108--3.

[32] Aliev AE, Lima MH, Silverman EM, Baughman RH. Thermal conductivity of multi-walled carbon nanotube sheets: radiation losses and quenching of phonon modes. Nanotechnology. 2010;21(3):035709.

[33] Shiomi J, Maruyama S. Molecular dynamics of diffusive-ballistic heat conduction in single-walled carbon nanotubes. Japanese Journal of Applied Physics. 2008;47(4R):2005. [34] Savin AV, Hu B, Kivshar YS. Thermal conductivity of single-walled carbon nanotubes. Physical Review B. 2009;80(19):195423.

[35] Thostenson ET, Ren Z, Chou T-W. Advances in the science and technology of carbon nanotubes and their composites: a review. Composites Science and Technology. 2001;61(13):1899-912.

[36] Ajayan PM, Schadler LS, Braun PV. Nanocomposite Science and Technology: Wiley; 2003.

[37] Tasis D. Carbon Nanotube-Polymer Composites: Royal Society of Chemistry; 2013. [38] Lin Y, Taylor S, Li H, Fernando KS, Qu L, Wang W, et al. Advances toward bioapplications of carbon nanotubes. Journal of Materials Chemistry. 2004;14(4):527-41. [39] Coleman JN, Khan U, Gun'ko YK. Mechanical reinforcement of polymers using carbon nanotubes. Advanced Materials. 2006;18(6):689-706.

[40] Fiedler B, Gojny FH, Wichmann MH, Nolte M, Schulte K. Fundamental aspects of nano-reinforced composites. Composites Science and Technology. 2006;66(16):3115-25. [41] Bauhofer W, Kovacs JZ. A review and analysis of electrical percolation in carbon nanotube polymer composites. Composites Science and Technology. 2009;69(10):1486-98. [42] Ma P-C, Siddiqui NA, Marom G, Kim J-K. Dispersion and functionalization of carbon nanotubes for polymer-based nanocomposites: A review. Composites Part A: Applied Science and Manufacturing. 2010;41(10):1345-67.

[43] Ma PC, Kim JK. Carbon Nanotubes for Polymer Reinforcement: Taylor & Francis; 2011.

[44] Lu K, Lago R, Chen Y, Green M, Harris P, Tsang S. Mechanical damage of carbon nanotubes by ultrasound. Carbon. 1996;34(6):814-6.

(42)

[45] Mukhopadhyay K, Dwivedi CD, Mathur GN. Conversion of carbon nanotubes to carbon nanofibers by sonication. Carbon. 2002;40(8):1373-6.

[46] Gojny FH, Wichmann MHG, Köpke U, Fiedler B, Schulte K. Carbon nanotube-reinforced epoxy-composites: enhanced stiffness and fracture toughness at low nanotube content. Composites Science and Technology. 2004;64(15):2363-71.

[47] Gojny FH, Wichmann MHG, Fiedler B, Bauhofer W, Schulte K. Influence of nano-modification on the mechanical and electrical properties of conventional fibre-reinforced composites. Composites Part A: Applied Science and Manufacturing. 2005;36(11):1525-35. [48] Gojny FH, Wichmann MHG, Fiedler B, Schulte K. Influence of different carbon nanotubes on the mechanical properties of epoxy matrix composites – A comparative study. Composites Science and Technology. 2005;65(15–16):2300-13.

[49] Gojny FH, Wichmann MHG, Fiedler B, Kinloch IA, Bauhofer W, Windle AH, et al. Evaluation and identification of electrical and thermal conduction mechanisms in carbon nanotube/epoxy composites. Polymer. 2006;47(6):2036-45.

[50] Thostenson ET, Chou T-W. Processing-structure-multi-functional property relationship in carbon nanotube/epoxy composites. Carbon. 2006;44(14):3022-9.

[51] Sandler JKW, Kirk JE, Kinloch IA, Shaffer MSP, Windle AH. Ultra-low electrical percolation threshold in carbon-nanotube-epoxy composites. Polymer. 2003;44(19):5893-9. [52] O'Connell MJ, Boul P, Ericson LM, Huffman C, Wang Y, Haroz E, et al. Reversible water-solubilization of single-walled carbon nanotubes by polymer wrapping. Chemical Physics Letters. 2001;342(3–4):265-71.

[53] Vaisman L, Wagner HD, Marom G. The role of surfactants in dispersion of carbon nanotubes. Advances in Colloid and Interface Science. 2006;128–130(0):37-46.

[54] O'Connell MJ, Boul P, Ericson LM, Huffman C, Wang Y, Haroz E, et al. Reversible water-solubilization of single-walled carbon nanotubes by polymer wrapping. Chemical Physics Letters. 2001;342(3):265-71.

[55] Negra FD, Meneghetti M, Menna E. Microwave‐Assisted Synthesis of a Soluble Single Wall Carbon Nanotube Derivative. Fullerenes, Nanotubes and Carbon Nanostructures. 2003;11(1):25-34.

[56] Tasis D, Tagmatarchis N, Bianco A, Prato M. Chemistry of carbon nanotubes. Chemical Reviews-Columbus. 2006;106(3):1105.

[57] Karousis N, Tagmatarchis N, Tasis D. Current Progress on the Chemical Modification of Carbon Nanotubes. Chemical Reviews. 2010;110(9):5366-97.

(43)

[58] Datsyuk V, Kalyva M, Papagelis K, Parthenios J, Tasis D, Siokou A, et al. Chemical oxidation of multiwalled carbon nanotubes. Carbon. 2008;46(6):833-40.

[59] Xing YC, Li L, Chusuei CC, Hull RV. Sonochemical oxidation of multiwalled carbon nanotubes. Langmuir. 2005;21(9):4185-90.

[60] Gao C, He H, Zhou L, Zheng X, Zhang Y. Scalable Functional Group Engineering of Carbon Nanotubes by Improved One-Step Nitrene Chemistry. Chem Mat. 2009;21(2):360-70.

[61] Felten A, Bittencourt C, Pireaux JJ, Van Lier G, Charlier JC. Radio-frequency plasma functionalization of carbon nanotubes surface O-2, NH3, and CF4 treatments. J Appl Phys. 2005;98(7).

[62] Sham ML, Li J, Ma PC, Kim JK. Cleaning and Functionalization of Polymer Surfaces and Nanoscale Carbon Fillers by UV/Ozone Treatment: A Review. J Compos Mater. 2009;43(14):1537-64.

[63] Asano K, Kondo D, Kawabata A, Takei F, Nihei M, Awano Y. Chemical modification of multiwalled carbon nanotubes by vacuum ultraviolet irradiation dry process. Japanese Journal of Applied Physics Part 1-Regular Papers Brief Communications & Review Papers. 2006;45(4B):3573-6.

(44)

CHAPTER 2:

OXIDATION BEHAVIOR OF MULTI-WALLED

CARBON NANOTUBES FLUIDIZED WITH OZONE

A paper published in ACS Applied Materials and Interfaces*

Danny C. Vennerberg a, Rafael L. Quirino a, Youngchan Jangb, and Michael R. Kessler a,c

a

Dept. of Materials Science and Engineering, Iowa State University, Ames, IA, United States b

Kumho Petrochemical R&BD Center, Daejeon, South Korea c

School of Mechanical and Materials Engineering, Washington State University, Pullman, WA

2.1 Abstract

Multi-walled carbon nanotubes (MWCNTs) were simultaneously fluidized and oxidized with gaseous ozone in a vertical reactor. Two different varieties of MWCNTs were compared to determine the versatility of the treatment and to elucidate the effect of defects on the oxidation behavior of MWCNTs. The extent of oxidation and nature of functional groups introduced on the nanotube surfaces were determined using Fourier-transform infrared spectroscopy (FTIR), X-ray photoelectron spectroscopy (XPS), and Boehm titration, and structural changes were monitored with Raman spectroscopy, scanning electron microscopy (SEM), and transmission electron microscopy (TEM). After only a few minutes of treatment, non-graphitic impurities were removed from the MWCNTs and significant levels of oxidation (~8 at% O) were achieved with very little damage to the nanotube sidewalls. Short O3 exposure resulted in primarily hydroxyl functionalities while longer exposure led to the

*

D. Vennerberg, R. L. Quirino, Y. Jang, M. R. Kessler: Oxidation Behavior of Multi-walled Carbon Nanotubes Fluidized with Ozone, ACS Applied Materials and Interfaces, 2014, 6, 1835-1842 (doi: 10.1021/am4048305).

(45)

formation of mainly carboxylic acid groups. Aliphatic defects present in the commercially-produced MWCNTs were found to play an important role in the oxidation mechanism. Because of its ability to remove impurities and evenly oxidize the sidewalls of nanotubes without the use of any solvents, the fluidized O3 reaction developed in this study was found to be an attractive option for industrial-scale MWCNT functionalization.

2.2 Introduction

Carbon nanotubes (CNTs) possess extraordinary mechanical, thermal, and electrical properties, making them ideal candidates for a wide range of applications in areas as diverse as energy storage,[1, 2] electronics,[3] and biosensors.[4, 5] CNTs are expected to be particularly suitable as reinforcement for composites due to their high aspect ratio. However, the addition of CNTs to polymers does not always engender outstanding mechanical properties and, in fact, has been reported to have deleterious effects in some instances.[6, 7] This under-performance stems from several possible sources. The graphene walls of as-grown CNTs are hydrophobic and have low surface energy, making CNTs incompatible with polar solvents and most polymer matrix materials. Furthermore, strong inter-tube van der Waals attractions cause CNTs to agglomerate into bundles, which have poor mechanical properties. Surface modification is an established approach to overcome poor dispersion and weak interfacial bonding in CNT-reinforced composites. Oxidation of CNTs by exposure to concentrated acids is a popular functionalization method at the laboratory-scale that has proven to be effective in improving composite properties.[8-12] However, acid exposure causes significant damage to the nanotube structure and leaves behind weakly adsorbed debris, which interferes with interfacial bonding in a composite if not removed.[13] Acid oxidation is also unattr

Development of scalable methods for the utilization of multi-walled carbon nanotubes in polymer and metal matrix composites

Development of scalable methods for the

Development of scalable methods for the utilization of multi-walled carbon nanotubes in polymer and metal matrix composites

TABLE OF CONTENTS

CHAPTER 1: GENERAL INTRODUCTION ...1

CHAPTER 3: EFFECT OF SILANE STRUCTURE ON THE PROPERTIES OF SILANIZED MULTIWALLED CARBON NANOTUBE-EPOXY

CHAPTER 5: ANISOTROPIC BUCKYPAPER THROUGH SHEAR-INDUCED MECHANICAL ALIGNMENT OF CARBON NANOTUBES IN WATER ...113

CHAPTER 7: GENERAL CONCLUSIONS...153

APPENDIX B: EFFECT OF FUNCTIONALIZATION ON THE RHEOLOGICAL BEHAVIOR OF MWCNT/EPOXY DISPERSIONS ... 190

LIST OF TABLES

LIST OF FIGURES

ACKNOWLEDGEMENTS

ABSTRACT

CHAPTER 1:

1.2 Dissertation Organization

Figure 1-1. SWCNT, DWCNT, and MWCNT structures (from left to right)[2]

Figure 1-2. Schematic showing how a graphene sheet (a) can be “rolled” up into zig-zag (b), armchair (c), and chiral (d) nanotubes[5]

Figure 1-3.Experimental setup used by Ruoff to perform tensile tests on individual CNTs[17]

Figure 1-4.Schematic depiction of noncovalent functionalization via polymer wrapping[42]

1.4 Research Objectives

1.5 References

CHAPTER 2:

2.1 Abstract

2.2 Introduction